How to pounce like a praying mantis

New high-speed video of young praying mantises leaping reveals how they manage to land cleanly every time.
09 March 2015

Interview with 

Professor Malcolm Burrows, University of Cambridge


How do falling cats always always manage to land on their feet? And why does toast tend to end up butter-side down? It's all down to how things spin. This is more important than it might sound- it could help us to build robots that can jump accurately from one place to another, something that's a major challenge right now. Cambridge University scientist Malcolm Burrows and his colleagues have been watching something else that is an expert at jumping- the praying mantis. Khalil Thirlaway went to meet him...

Malcolm - In their natural habitat, they live amongst the stems of plants or the branches of trees. When they're young, they don't have any wings. So, if they're on these branches, the only way to get to another branch basically is to jump. So, we arranged them to stand on a platform and gave them this small black rod that's a target to which they could jump. They would jump reliably time after time to this target. And that enabled us to then film these jumps about a thousand frames a second and that enabled us to slow the time down sufficiently so that we could resolve the intricate movements that they were making during this very rapid movement.

Khalil - So, how are they generating this power, also with this accuracy of landing every single time?

Malcolm - First of all, what they do is they sit on their platform and they scan the target by moving their head backwards and forwards seemingly to estimate the distance that they've got to jump. Then they prepare themselves for jumping, adjusting the centre of mass of their body by curling their abdomen over the top of the body so that when the middle legs and the hind legs provide the propulsive force for a jump, there is a spin imparted to the body as it takes off. Once they're airborne, they have to regulate the amount of spin that the body is going through so that it's correctly oriented at the right angle to be able to land with all their legs and grab hold of the target. So, how do they control this spin? Once they're airborne, there's nothing to push against except the air. They don't have wings. Well, they go for this intricate sequence of movements in which they rotate the front legs, the hind legs and the abdomen in different directions and in different combinations throughout the jump. This rotation of these limbs and the abdomen stores angular momentum and means that the body itself doesn't rotate as much and remains stable and in the correct orientation for landing on the target.

Khalil - How do we know that these movements are actually fine-tuning the mantis's angle?

Malcolm - We have two approaches to doing that. first of all, we built a model. Not an artificial model of a mantis, but a mathematical model of what the body is doing. And then we could say, well, what happens if one of these parts of the body does not move? First thing we did was to say, "Well, what happens if the abdomen does not curl up?" the answer to that was, that it didn't impart the necessary spin onto the body. What in fact happened without the abdominal rotation was that it went headfirst into the target and just head butted the target and then sort of slipped off.

Khalil - That's not quite as graceful as the previous scenario.

Malcolm - Not at all and that was just messing up with one part of the body. And it told us a lot about the mechanisms that were underlying these movements. The second thing was to experimentally interfere with the movements. So, one of the experiments we did was to glue the segments of the abdomen together so that it couldn't curl up as much. Again, what happened there was that it crashed into the target because it couldn't rotate its body enough.

Khalil - Now we understand the mechanics of how the mantis fine-tunes its orientation in the jump, how do you think this could be applied in other areas of research?

Malcolm - Well, I think it has intrinsic value of its own to start with. This is a very complicated movement and the mantis has come up with some intricate mechanisms of how to solve the issue. But the application could be in robotics. One of the things that roboticists are trying to do is to build robots that can actually jump. One of the huge design problems that they face in doing that is that once the robot has left the ground, it tends to spin uncontrollably.

Khalil - So, it seems that jumping isn't the problem. It's getting robots to land safely which is the big problem.

Malcolm - That's right. it's getting them to be stable in the air. So, taking some of the mechanisms that the mantis use is, should enable roboticists to programme their robots much more reliably and give them stability once they're in the air.

Khalil - This seems like, yet, another example of taking inspiration from nature which just had billions of years to evolve and applying it to our problems.

Malcolm - Well that's right. we've only been trying to design things ourselves for a few tens of hundreds of years. So, why not take advantage of what's gone before us, billions of years of history providing solutions that may be of use in our designs.


Add a comment